Impact of Injuries on Team Performace

[AcGold]

this won't end well

[Poe969]

Originally Posted by Cali Flames Fan

I know you're being a jerk here, but now I'm actually intrigued. I'm not a fan of his analysis at all, but here's a place it could be considered useful…possibly projecting man games lost due to style of play.

It should follow that a player with a higher GRIT index as per Ricardo's analysis should be injured more often because they put their body at risk more often. I challenge Ricardo to put together a massive data analysis project together and find whatever correlations he can between his GRIT index and being prone to injury. Might prove interesting.

Kind of being a jerk kind of not.... The whole GRIT thing is a stretch but it is also kind of valid. The question here is; do players with more grit get injured more often and is it worth having someone who plays with more grit for a part of the season or someone with less grit for more of the season.

Guys like Glencross get injured more often because he played a certain style and now his body can't keep up. Darcy Tucker was another example of players that play "all out" and then miss time because of the way they play.

[Lanny_McDonald]

Originally Posted by ricardodw

Hey PeteMoss ... a post like this will generate a almost visceral response from a segment of CP.

I posted the Flames personal win % in this thread link

The immediate response of my attempt of correlating team success to individual contribution was:




Thanked by Cheerio, hummdeedoo, New Era, Resolute 14, SvenTastic47, TorqueDog, V, Wastedyouth, wireframe..

I am looking for some valid hockey knowledgeable post from this group of experts.

I'll take the bait.

No, what you are looking for is validation of your obsession in trying to model hockey using fudged statistics. You coming on here and telling us that injuries happen and the better the player the bigger the impact to the team they play for is not earth shattering news. We've known this for decades. You trying to show it with some garbage model dreamed up in your basement is just more validation to the theory from the thread you pointed to above; you come up with a theory and then try to find statistics to support it rather than actually analyzing data to find significant trends. Your observations here are not ground breaking. It is like taking accident information and concluding that children who play in traffic are more likely to be hit cars. No kidding??? Hockey players that play a more physical brand of hockey are going to get hurt a lot more. Players who block a lot of shots are going to get hurt a lot more. Players who put themselves into situations where they take or initiate contact are going to be hurt more. Teams that suffer injuries to their best players are going to feel those injuries more than teams that suffer injuries to their support players. Those are all well known facts. They are also common sense. We didn't need a Ricardo Injury Index to tell us something we already knew.

[ricardodw]

Teams with 6 important players missing: Pittsburgh, Calgary
5: Detroit
4: Winnipeg, Minnesota, Boston
3: Montreal, NJ, NYI, Vancouver

This year these 9 teams

Pittsburgh #4, Flames #8, Detroit #12, Winnipeg #15, Minnesota #18, Boston #11, Montreal #1, NJ #20, NYI #9 , Vancouver #7

Flames, Detroit, Montreal, NYI, Vancouver, Winnipeg 6 of these teams are doing significantly better than the pre - season analysts had them.


The Flames are doing a lot better and there is a strong argument imho that adding Wideman #3, Glencross #8 Giordano #13 and Russell #34 most injured impact players is the reason for their vastly improved results.

The Jets are a different team with Bogosian, Trouba and Kane in their line-up

[MisterJoji]

Originally Posted by ricardodw

Teams with 6 important players missing: Pittsburgh, Calgary
5: Detroit
4: Winnipeg, Minnesota, Boston
3: Montreal, NJ, NYI, Vancouver

This year these 9 teams

Pittsburgh #4, Flames #8, Detroit #12, Winnipeg #15, Minnesota #18, Boston #11, Montreal #1, NJ #20, NYI #9 , Vancouver #7

Flames, Detroit, Montreal, NYI, Vancouver, Winnipeg 6 of these teams are doing significantly better than the pre - season analysts had them.


The Flames are doing a lot better and there is a strong argument imho that adding Wideman #3, Glencross #8 Giordano #13 and Russell #34 most injured impact players is the reason for their vastly improved results.

The Jets are a different team with Bogosian, Trouba and Kane in their line-up

[Party Elephant]

New Era's post was a mic drop that shut this thread down for 2 whole months.

[DeluxeMoustache]

I have no issues with a guy interested in hockey and the Flames stepping outside of the box and finding data to support his hypotheses.

Yes, key players get hurt on many teams but not equally so.

Having said that, the ability of the guys stepping in to contribute and mitigate the losses will be another factor entirely.

The Flames are impressing a lot of people this year and some would argue that last year was impressive in terms of on ice product but not so much in terms of results. Heck, even Aaron Ward predicted playoff potential here this year.

Thumbs up Ricardodw, stay the course.

[Jay Random]

I have every 'issue' with people who spend their time finding data to support their hypotheses.

They're doing it wrong. In fact, most of the criminal stupidity in human history has been caused by people thinking that way. (Fortunately, stupidity about hockey isn't important enough to count as a crime.)

If you want to do something useful, try to look at the data first, and then come up with hypotheses that explain it.

[ricardodw]

Originally Posted by Jay Random

I have every 'issue' with people who spend their time finding data to support their hypotheses.

They're doing it wrong. In fact, most of the criminal stupidity in human history has been caused by people thinking that way. (Fortunately, stupidity about hockey isn't important enough to count as a crime.)

If you want to do something useful, try to look at the data first, and then come up with hypotheses that explain it.

Hey Jay

I looked at the data and came up with a hypothesis that teams that had significant injuries in a given year had worse years than what their talent base would normally provide. This was a factor that would be corrected provided they reverted back to the median/ average injury.


The teams that had significant injuries last year did not meet expectations. As a result the vast majority of punditry were unable to overcome the the lower expectations brought on by the randomness of the injury situation.

They picked the Flames to be a bottom 3 team.


Had Wideman, Glencross, Gio and Russell all been signed by any random team without giving up any existing assets then it would not any great surprise that they turned out significantly better than they were the previous year.


Likewise had the Flames signed Trouba, Bogosian and Kane to add to last years lineup there would have been playoff expectations coming out of training camp.


So far...with limited data... after 1/4 of a season my hypothesis has been correct in 6 out of 9 cases.... and wrong in 2 (NJ , Min) out of 9 cases.

Jay... you might want to pause and reflect that your self proclaimed hockey genius might not extend beyond your own somewhat limited mind set.

[Jay Random]

Originally Posted by ricardodw

Jay... you might want to pause and reflect that your self proclaimed hockey genius might not extend beyond your own somewhat limited mind set.

What 'self-proclaimed hockey genius' is this? When have I ever proclaimed myself a genius? Where? Quote or retract.

As for my 'somewhat limited mind set', the fact that I disagree with you does not prove that I have a limited mind. Most people on here disagree with you; does that mean that you are the only person on CP who is not a fool?

[Street Pharmacist]

Originally Posted by Jay Random

I have every 'issue' with people who spend their time finding data to support their hypotheses.

They're doing it wrong. In fact, most of the criminal stupidity in human history has been caused by people thinking that way. (Fortunately, stupidity about hockey isn't important enough to count as a crime.)

If you want to do something useful, try to look at the data first, and then come up with hypotheses that explain it.

????

That's how science works:

Hypothesis first, look for data to test Hypothesis


You can't and shouldn't do it backwards

[Jay Random]

Originally Posted by Street Pharmacist

????

That's how science works:

Hypothesis first, look for data to test Hypothesis


You can't and shouldn't do it backwards

No. That is not what is at issue here.

What I am complaining about is this process:

1. Make hypothesis.
2. Look for data that supports hypothesis, and ignore data that falsifies it.

A hypothesis that accounts for only some of the data, but is contradicted by the body of data as a whole, is not only unscientific, but dangerous. It is worse than worthless.

As for making the hypothesis before looking for data, that's wrong too. The whole point of making hypotheses in the sciences is to try to come up with explanations for observed phenomena. You have to start with the phenomena that you wish to explain, and that can only be done by observation.

[codynw]

Injuries to good players? Still a bad thing. Shocking, right?

[Bandwagon In Flames]

How is this a thread? Why is it necessary to provide data for something that is quite obviously a bad thing.

Good players injured and depth at position = NOT good for team, but team can still win games
Good players injured and no depth at position = NOT good for team, and team will likely plummet in the standings
Good players injured = NOT good for team

Ricardo, while you may actually have valid and relevant data to support this, you do a terrible job of conveying this to others. So many long posts that ramble on about 20 different arguments and then end with a conclusion that's barely relevant to your different sources of data.



"I have every 'issue' with people who spend their time finding data to support their hypotheses."

Jay, I think you have a point, but need to re-word this unless you hate scientists and such. I do agree that it's tiresome to read posters grasping at straws bringing up irrelevant data to support their claims.

[Street Pharmacist]

Originally Posted by Jay Random

No. That is not what is at issue here.

What I am complaining about is this process:

1. Make hypothesis.
2. Look for data that supports hypothesis, and ignore data that falsifies it.

A hypothesis that accounts for only some of the data, but is contradicted by the body of data as a whole, is not only unscientific, but dangerous. It is worse than worthless.

As for making the hypothesis before looking for data, that's wrong too. The whole point of making hypotheses in the sciences is to try to come up with explanations for observed phenomena. You have to start with the phenomena that you wish to explain, and that can only be done by observation.

That's not what at all what you said though. You completely changed your stance here. Plus, what you described is exactly what Ricardodw did. He thought it might have an impact. He sought to quantify it by looking at data. Results posted

Why do people jump on him for that? I get he's a polarizing poster and some like to jump on every word to make themselves feel better, but I'm unclear what is wrong with what he did here

[Jay Random]

Originally Posted by Bandwagon In Flames

"I have every 'issue' with people who spend their time finding data to support their hypotheses."

Jay, I think you have a point, but need to re-word this unless you hate scientists and such. I do agree that it's tiresome to read posters grasping at straws bringing up irrelevant data to support their claims.

I don't 'hate scientists and such' and haven't said anything to imply it.

A good scientist is far more concerned about finding data that might falsify his hypothesis, because that will tell him when the hypothesis needs to be changed. Cherrypicking data to support your predetermined conclusion is the very opposite of scientific.

The gold standard here is exemplified by Einstein's attitude towards his own theory of general relativity. His theory predicted three effects that should be observable in the behaviour of light waves. Two of those effects were observed by studying solar eclipses, but the third (the gravitational red shift) was not detected until some years later. Einstein knew that two out of three was not good enough, and after the eclipse studies he said about the red shift:

'If it were proved that this effect does not exist in nature, then the whole theory would have to be abandoned.'

I hold out the highest hopes for advanced stats in hockey, though I do think their value will be limited as long as they measure proxies instead of actual events. But these hopes are a long way from being fulfilled. Too many self-proclaimed advanced stats guys are behaving like a little kid with a shiny new hammer, running around hitting random things in the hope that some of them will turn out to be nails.

[Jay Random]

Originally Posted by Street Pharmacist

That's not what at all what you said though. You completely changed your stance here.

No, I did not change my stance at all. You misunderstood me the first time, so I tried to state my point more clearly.